CentraleSupélec (3, rue Joliot Curie,
Plateau de Moulon,
91192 GIF-SUR-YVETTE Cedex - France)
Abstract : In this paper, we determine the analytical form of the interaction matrix related to any moment that can be computed from segmented images. The derivation method we present is based on Green's theorem. We apply this general result to classical geometrical primitives. We then consider using moments in image-based visual servoing. For that, we select six combinations of moments to control the six degrees of freedom of the system. These features are particularly adequate if we consider a planar object and the configurations such that the object and camera planes are parallel at the desired position. The experimental results we present show that a correct behavior of the system is obtained if we consider either a simple symmetrical object or a planar object with complex and unknown shape.
https://hal.inria.fr/inria-00352019
Contributor : Eric Marchand <>
Submitted on : Monday, January 12, 2009 - 1:52:19 PM Last modification on : Friday, January 8, 2021 - 3:12:44 AM Long-term archiving on: : Tuesday, June 8, 2010 - 7:30:10 PM
François Chaumette. Image moments: a general and useful set of features for visual servoing. IEEE Transactions on Robotics, IEEE, 2004, 20 (4), pp.713-723. ⟨inria-00352019⟩